R1 / R2 traceability model
Purpose. Answer six questions without archaeology: refer date and amount, recall date and balance, reasons, DCA-reported snapshot, and what was proposed versus what CP actually did next. Implement as columns in the CP recoveries book or equivalent, fed by placement, recall, and DCA state files.
Audit questions this must answer
- When was this account referred to DCA1, and at what refer amount?
- When was it recalled from DCA1, at what balance, and why?
- What did the DCA believe the balance and arrangement status were at recall?
- Why did CP move it to DCA2 (strategy), as distinct from “DCA asked”?
- Same four questions for R2 if used.
R1 field set (illustrative names)
| Field | Description |
|---|---|
| R1_DCA | Identifier of first DCA vendor |
| R1_Refer_Date | CP-recognised placement date |
| R1_Refer_Amount | Authoritative refer balance at placement |
| R1_Current_Balance | CP authoritative balance while at R1 (rolling) |
| R1_Status | OPEN / RECALL_PENDING / CLOSED |
| R1_Recall_Request_Date | When DCA sent RQT if applicable |
| R1_DCA_Reported_Recall_RQT_Date | As reported by DCA (should match R1_Recall_Request_Date when aligned) |
| R1_DCA_Reported_Recall_RQT_Reason | Vendor reason text or code for the request |
| R1_Recall_Date | When CP executed stop work (actual recall) |
| R1_Recall_Reason | CP-coded reason (compliance, strategy, customer paid, and so on) |
| R1_Recall_Category | COMPLIANCE | STRATEGIC | DATA_CONTROL (taxonomy yours) |
| R1_Recall_Balance | CP balance at recall event |
| R1_DCA_Reported_Balance | Snapshot from DCA state file near recall |
| R1_DCA_Reported_Status | DCA workflow status |
| R1_DCA_Reported_Arrangement_Status | Active, broken, none, and so on |
| R1_Next_Strategy_Proposed | What DCA or ops proposed (R2, internal, sale) |
| R1_Next_Strategy_Actual | What CP approved (must be populated before R2 refer if applicable) |
These fields exist because recall request and actual recall are different moments; category explains regulator and audit questions; balance at recall anchors cohort and fee arguments; proposed versus actual next strategy explains DCA1 to DCA2 moves without relying on oral history.
R2 field set
Mirror the same pattern with R2_ prefixes. R2 must never be “empty history.” If R2 exists, R1 must be closed with recall timestamps and balances.
Recall request versus actual recall
RQT is not permission by itself. Store both the request timestamp and the CP decision (approve or reject) and the actual recall execution time. SLAs often require stop work within hours for conduct cases; strategy recalls may batch daily. Reporting should not confuse “asked” with “stopped.”
Proposed versus actual next strategy
DCA may propose R2 with commercial motivation. CP may choose internal work instead. If you only log “recalled,” you lose the decision quality story for NPV feedback and vendor scorecards.
Why this supports reporting
Vintage and spin-down reports need clean cohort boundaries. If R1 and R2 placements overlap in reporting because recall was late, recovery rate numerators double-count or miss cash. Trace fields align cohort membership to CP truth, not DCA campaign labels.
Why this supports DCA comparison
Comparing DCA vendors fairly requires similar populations and handoff points. If one vendor receives more “cold” R2 transfers from failing peers, raw recovery curves mislead unless reasons and refer balances are visible.
Related: CP / DCA files · Reporting data lineage · Back to pack home